An Assessment of the Toulouse Saint Louis University Mini Falls Assessment Tool to Predict Incident Falls among Older Adults Residing in Nursing Homes: A 6-Month Prospective Study

OBJECTIVES: Toulouse Saint Louis University Mini Falls
Assessment (TSLUMFA) tool has been designed to predict falls.
It was initially validated in a geriatric clinic in 2018. The primary
objective was to evaluate the predictive capacity of the TSLUMFA for
incident falls in older adults residing in nursing homes. The secondary
objective was to determine the TSLUMFA optimal cut-off value
identifying those older adults with a high-risk of falling.
SETTINGS: A longitudinal study was carried out over a period of six
months.
PARTICIPANTS: 93 older adults residing in nursing homes were
evaluated for the present study.
MEASUREMENTS: The TSLUMFA (made up of 7 criteria) was
administered at baseline, and incident falls were recorded based on a
registry of falls. Comparisons of TSLUMFA scores between fallers
and non-fallers were performed using the U Mann-Whitney test or
Chi². Correlation between the total TSLUMFA score (/30 points) and
incident fall(s) was explored using the Cox proportional hazard model.
ROC analysis enabled an optimal cut-off value to be established to
identify those adults at the highest-risk of falling.
RESULTS: In the study, 93 older adults (61.3% women) with a median
age of 80 (69-87) years were included. The median total TSLUMFA
score was 21 (19-24.5) points. During the 6-month study period, 38
subjects (40.9%) experienced at least one fall. The total TSLUMFA
score in older adults with incident fall(s) was significantly lower than
in those who did not fall (20 (15.75-22.25) points versus 23 (20-25)
points and a p-value of <0.001). For each 1-point higher score at the
total TSLUMFA a 9% less chance of falling was observed during
the study period (p-value = 0.006). The AUC was 0.736 (95%CI:
0.617-0.822) and p-value <0.001, clearly demonstrating its interesting
performance as a screening tool. A score of ≤ 21 points was identified
as the optimal cut-off to identify those older adults at a higher-risk of
falling.
CONCLUSION: The TSLUMFA performed well and successfully
identified older adults with a high risk of falling in a nursing home
setting. Further comparisons with existing tools are warranted.